# AlayaWorld: Interactive Long-Horizon World Modeling - Full Technical Report (v1.1)

arXiv: https://arxiv.org/abs/2608.13492 · HTML: https://mlsanity.com/paper/2608.13492/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Placed in Video-generation world models and simulators, one of 11 directions in World models.
- Builds on 3 of its 3 references that are in the map.
- First posted on arXiv 13 Aug 2026.

## Placed in

- World models / [Video-generation world models and simulators](https://mlsanity.com/wm/video-generation-world-models-and-simulators/)

## Builds on (most cited first, 3 of 3)

- [Depth Anything 3: Recovering the Visual Space from Any Views](https://mlsanity.com/paper/2511.10647/) (2025)
- [WBench: A Comprehensive Multi-turn Benchmark for Interactive Video World Model Evaluation](https://mlsanity.com/paper/2605.25874/) (2026)
- [Towards Consistent Video Geometry Estimation](https://mlsanity.com/paper/2605.30060/) (2026)

## Built on by (most cited first, 7 of 7)

- [SolarWM: Open Data and Scalable Training for Long-Horizon Video World Models](https://mlsanity.com/paper/2609.02886/) (2026)
- [AlayaVista: Streaming World Modeling from Panoramic States to Perspective Video](https://mlsanity.com/paper/2609.14462/) (2026)
- [AI for Games in the Foundation Model Era](https://mlsanity.com/paper/2609.16679/) (2026)
- [Code Plans, Diffusion Renders: Open-Ended Generative World Modeling](https://mlsanity.com/paper/2609.26458/) (2026)
- [MUGEN: Interactive Panoramic World Exploration via Camera Control](https://mlsanity.com/paper/2609.38077/) (2026)
- [LOCI: Spatial Linear Memory for Streaming World Models](https://mlsanity.com/paper/2609.40222/) (2026)
- [RIGen: Beyond Action Following with Reciprocal Interaction in Generative Gameplay](https://mlsanity.com/paper/2610.20811/) (2026)

## Abstract (arXiv)

This report presents an improved version of AlayaWorld. While the backbone architecture, chunk-wise autoregressive generation scheme, and training data remain unchanged from the previous release, we substantially revise how conditioning signals are represented and integrated into the model. The new design is guided by a simple principle: conditioning signals should match the generated content as closely as possible in both latent representation and temporal structure. To this end, we make two major changes. First, we replace the previous depth-warping-based spatial memory with a streaming 3D point-cache renderer. Second, we redesign the conditioning pipeline so that visual conditions are encoded in the same causal-VAE latent space, with temporal statistics consistent with those of the generated video. Concretely, the new version introduces six modifications: (1) replacing static-frame image conditioning with motion-aware latent conditioning; (2) causally encoding re-rendered spatial memory as a continuous sequence; (3) aligning the temporal-memory window in pixel space; (4) adopting hard memory dropout that removes memory tokens rather than zeroing them; (5) unifying the VAE encoding and decoding protocol across training and inference; and (6) removing the camera AdaLN branch, such that viewpoint control is provided entirely through the re-rendered spatial condition.